Thanks to visit codestin.com
Credit goes to www.scribd.com

Scribd
Upload
96 views8 pages

LQR Control Design Guide

The document discusses linear quadratic regulator (LQR) control. It defines the LQR problem as designing an optimal controller for a linear system where the cost is quadratic. The solution is a state feedback controller defined by the LQR gain matrix K. It describes the process of designing the LQR controller by choosing K to minimize a quadratic cost function subject to the system dynamics, and solving the resulting algebraic Riccati equation to determine K.

Uploaded by

Khoa Nguyễn
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
96 views8 pages

LQR Control Design Guide

The document discusses linear quadratic regulator (LQR) control. It defines the LQR problem as designing an optimal controller for a linear system where the cost is quadratic. The solution is a state feedback controller defined by the LQR gain matrix K. It describes the process of designing the LQR controller by choosing K to minimize a quadratic cost function subject to the system dynamics, and solving the resulting algebraic Riccati equation to determine K.

Uploaded by

Khoa Nguyễn
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 8

Vietnam National University Ho Chi Minh City

HCMC University of Technology


Faculty of Mechanics Engineering
Department of Mechatronics Engineering

Subject:
Linear System and Control

LQR Control

TS. Ngô Hà Quang Thịnh, [email protected]


Linear Quadratic Regulator
Definition of LQR
The theory of optimal control is concerned with operating a dynamic
system at minimum cost. The case where the system dynamics are
described by a set of linear differential equations and the cost is
described by a quadratic function is called the LQ problem. One of the
main results in the theory is that the solution is provided by the linear–
quadratic regulator (LQR), a feedback controller whose equations are
given below:

Q, R is the positive (semi-) definite matrix

TS. Ngô Hà Quang Thịnh, [email protected]


Design of LQR Controller
By choosing the state feedback control rule u = - Kx with K is a tuning
gain, then substituting into equation of J
1 T
J =  x (Q + K T RK ) xdt
20
In order to compute K, it could be defined from Lyapunov function.

Lyapunov candidate function is


1 T 1
V ( x (t )) =  x (Q + K T RK ) xdt = x T Px
2t 2
V(x(0)) = J = xT(0)Px(0)
Derivating the first order
1
V ( x ) = x T (Q + K T RK ) x |t
2
1
  1
 
= x T ( ) Q + K T RK x ( ) − x T (t ) Q + K T RK x (t )
2 2

TS. Ngô Hà Quang Thịnh, [email protected]


Design of LQR Controller
Assume that selecting K in order that x() →0
1
 
V ( x ) = − x T (t ) Q + K T RK x (t )
2
Furthermore,


2
T 1 T
2

V ( x) = (x Px + x Px ) = x ( A − BK )T P + P ( A − BK ) x
1 T

Then
1 T
2
 1 T
x ( A − BK ) P + P ( A − BK ) x = − x (Q + K T RK ) x
T
2
Matrix P satisfy Lyapunov function
( A − BK )T P + P ( A − BK ) = −(Q + K T RK )

TS. Ngô Hà Quang Thịnh, [email protected]


Design of LQR Controller
Step 1: Solving the Lyapunov candidate function to obtain matrix P in
according with matrix K
1
Step 2: Then compute J = V(x(0)) = xT (0) Px (0) is a function of matrix K
2
J P
Step 3: To get min of J, we solve = 0 or =0
kij kij
Step 4: Obtain matrix K, the control rule u = - Kx

Step 5: Consider the stability of (A –BK)

If the outputs are adjustable, then choose


1 T T
J =  x (C QC + K T RK ) xdt
20
With R = T ,  is square matrix

TS. Ngô Hà Quang Thịnh, [email protected]


Design of LQR Controller
Lyapunov function can be re-written

( AT − K T B T ) P + P ( A − BK ) + Q + K T T K = 0
AT P + PA + [K − (T ) −1 B T P ]T [K − (T ) −1 B T P ] − PBR −1 B T P + Q = 0
P
Derivating with kij and using =0
kij
We have

[(K − (T ) −1 BT P )T (K − (T ) −1 BT P )] = 0
kij
The optimization could be achieved if
K = (T ) −1 B T P
K =  −1 (T ) −1 B T P = R −1 B T P
In this case, Lyapunov function becomes algebraic equation Riccati
AT P + PA − PBR −1 B T P + Q = 0

TS. Ngô Hà Quang Thịnh, [email protected]


The End

You might also like